Williams' absolute conception

Later today I’m teaching a session on Williams’ absolute conception in my first year Value of Knowledge module (BA Philosophy). I realise that I don’t or no longer understand a key point. If I work through the argument (below) as I've previously thought about it, perhaps this will help pinpoint my blindspot.

Williams’ way of substantiating the Absolute Conception develops from a minimal thought experiment. Two people, A and B, both have different but equally true beliefs about the world. (Perhaps they are in different places). If so, then it must be possible to describe the set up, including how A and B’s beliefs, as different true representations of the same broader thing, are related both to each other and to the thing of which their respective beliefs are representations. But this larger account is itself a representation of the world that includes A and B and explains how their beliefs are related to each other and to the world.

Hence, we have formed a conception of a larger representation of the part of the world in question (now construed broadly to include both A and B and their original representations), which is distinct from our conception of that part of the world itself. Thus understanding the larger story requires also grasping the contrast between, on the one hand, the representation of the world, and, on the other, the world of which the representation is (merely) a (albeit true) representation.

In fact postulating two people seems unnecessary for the thought experiment as indicated by the fact that the last ‘larger story’ / representation can itself be the subject of a broader account of how it itself relates to, but is distinct from, the world it represents and to other true representations of the world that could be given. And so in earning the right to the initial aspect of knowledge – that it is of something independent – we seem to have a kind of escalation of representations, each one standing distinct from the complex part of the world it represents.

This makes me think that there’s a quicker route to the same idea starting from Williams’ premiss: ‘if knowledge is what it claims to be, then it is knowledge of a reality which exists independently of that knowledge, and indeed... independently of any thought of experience’ [Williams 1978: 64]

So, with Williams, using the word ‘representation’ to mean anything which represents the world or reality as being a particular way, we can say: if our conception of knowledge is of something independent of us, we must also think of the knowledge in question (or the representation) as distinct from, but depicting, a bit of the world. And thus our everyday understanding of knowledge must involve a contrast between:

• A: a representation, on the one hand, and the world, on the other.

Williams argues that the very idea of knowledge involves that contrast. Anyone who did not understand the idea that, although knowledge represents the world being a particular way, it is distinct from the world, would not understand what knowledge is. But that point - ‘A’ - is itself expressed in a representation which, because it aims at objective truth, stands in contrast to another bit of reality (a bit of reality which also includes a representation). And so to grasp the objectivity of A requires grasping this further contrast between:

• B: a representation (ie A: of a representation, on the one hand, and the world, on the other), on the one hand, and the world (what A represents: a more complicated bit of the world this time which also includes the first representation), on the other

But ‘B’ is itself something that we think and take to be true and so is another representation. So it is distinct from the state of affairs it depicts. So to grasp its objectivity - its aim to tell us something about how true representations are nevertheless distinct from the world they represent - we must be capable of thinking about a further distinction which we can label ‘C’. And so on. This line of thinking is suggested by this passage:

For if A or B or some other party comes in this way to understand these representations and their relation to the world, this will be because he has given them a place in some more inclusive representation; but this will still itself be a representation, involving its own beliefs, conceptualizations, perceptual experiences and assumptions about the laws of nature. If this is knowledge, then we must be able to form the conception, once more, of how this would be related to some other representation which might, equally, claim to be knowledge; indeed we must be able to form that conception with regard to every other representation which might make that claim. If we cannot form that conception, then it seems that we do not have any adequate conception of the reality which is there ‘anyway’, the object of any representation which is knowledge; but that conception appeared at the beginning as basic to the notion of knowledge itself. That conception we might call the absolute conception of reality. If knowledge is possible at all, it now seems, the absolute conception must be possible too. [Williams 1978: 65]

So much for the argument as to why we must be able to form an absolute conception of the world if we are to grasp the concept of knowledge. But how we do so faces a dilemma:

On the one hand, the absolute conception might be regarded as entirely empty, specified only as ‘whatever it is that these representations represent’. In this case, it no longer does the work that was expected of it, and provides insufficient substance to the conception of an independent reality; it slips out of the picture, leaving us only with a variety of possible representations to be measured against each other, with nothing to mediate them.
On the other hand, we may have some determinate picture of what the world is like independent of any knowledge or representation in thought; but then that is open to the reflection, once more, that that is only one particular representation of it, our own, and that we have no independent point of leverage for raising this into the absolute representation of reality. [ibid: 65]

Our concept of knowledge seems to require a contrast between representations and the world represented. Awareness of this seems to escalate into a conception of what all possible true representations represent: an absolute conception. But

• Either it is just, vacuously, stated by saying ‘what true representations represent’.
• Or we use a particular picture of the world. But in our own case, that just trades on some particular representation. It may, eg., be false.

And hence we need some other response to this which turns out to be an indirect way of specifying the content (in one sense) of the conception via a Piercean ideal endpoint of inquiry route: reality is what is described by a finished science. We can conceive reality to the extent that we can concieve of this process towards an endpoint. Now as McDowell points out, pretty much the same kind of dilemma we had earlier arises again with this suggestion in characterising science (either ‘whatever gives us access to reality’ or that historical method practiced in such and such a lab). But what bothers me today is getting a feel for the dilemma in the first place. What exactly is the worry about the first horn?

The minimal thought experiment suggests that whenever we have a true representation (the sort of thing that knowledge is, at the very least) it is possible to tell a broader story of how it relates to the reality, a bit of which it depicts, and that this might go on without stopping short of an account of how the totality of true representations depict the totality of reality / the world as a whole. But to get as far as the dilemma requires more than just the thought that it might be possible to escalate (that there need be no stopping point). Rather, it requires thinking that we need to escalate in this way to earn the right to thinking of knowledge as answering to an independent reality in the first place. Ie only by doing this do we fill out our grasp of the world to which knowledge / representations answer.

With that worry in place then I think I see why the first horn is a problem. All the subsequent representations seem increasingly to concern representations rather than the world in contrast to which the first representation stood. Or, if it isn’t quite that, then if there was a problem understanding the objective purport of the first one, then embedding it in a series of further representations makes no progress. That idea was, I think, sustained for me in the past by adding sotto voce the phrase ‘the hell’ after ‘whatever’ in the first horn: whatever the hell it is that these representations represent. With that in place, the first horn seems to be a case of simply failing to specify a world to contrast to our representations.

But the glaring problem with that reading is that it makes the first step so odd. (I have to say that I always thought that the first step was the misleading one. What has changed is that I no longer understand its initial appeal.) Under what circumstance would one grasp something about the nature of knowledge or representation but lack a grasp of the bit of reality it aimed to get right that would, in turn, be informed or corrected by the first step of the escalation? Especially given that Williams is concerned with knowledge and thus true representations, what would it be to understand the true representation ‘Sootica is in her basket’ and not grasp that it is the fact that Sootica is in her basket that this aims to get, and succeeds in getting, right?